Researchers from the Wind Energy Institute at the Technical University of Denmark have developed a new data-driven approach to model wind farm wake flow, which could significantly improve wind farm layout and power forecasting.
Wind farm operators and planners have long struggled with accurately predicting wind flow and the wake effects that can significantly reduce power output. Traditional methods like physical measurements and numerical simulations can be time-consuming and resource-intensive. Physics-based models, while faster, can lack accuracy due to oversimplified physics. Data-driven models, which use large amounts of data to identify patterns and make predictions, have gained popularity in recent years.
The Danish researchers propose a new data-driven model inspired by video-frame interpolation and the principle of similarity. The method transforms field data into images and uses multi-scale feature recognition to identify, match, and interpolate wake structures. This is done using Scale-Invariant Feature Transform (SIFT) and Dynamic Time Warping (DTW) algorithms, which generate intermediate flow fields.
The researchers validated their approach using six representative mini wind-farm cases, which included variations in turbine spacing, size, combined spacing-size variations, different turbine counts, and wind-direction misalignment. The method achieved a mean absolute percentage error (MAPE) of 0.68-2.28% across all cases.
One of the key advantages of this new method is its computational efficiency. It can flexibly compute both 2D and 3D wake fields, offering substantial gains over large-eddy simulation (LES) and Meteodyn WT when 2D accuracy suffices for industrial needs. This makes it a practical alternative to measurements, high-fidelity simulations, and simplified physics-based models.
The researchers suggest that their method could enable efficient expansion of wake-flow databases for wind-farm design and power prediction, balancing speed and accuracy. This could lead to more efficient wind farm layouts and improved power forecasting, ultimately increasing the profitability and reliability of wind energy.
The research was published in the journal Renewable Energy.
Practical applications for the energy sector include:
1. **Wind Farm Layout Optimization**: By providing accurate and efficient predictions of wind flow and wake effects, this method can help optimize wind farm layouts to maximize power output.
2. **Power Forecasting**: Improved wake-flow modeling can enhance power forecasting, allowing grid operators to better integrate wind energy into the grid and maintain system stability.
3. **Reduced Operational Costs**: By reducing the need for time-consuming and resource-intensive physical measurements and numerical simulations, this method can help lower operational costs for wind farm operators.
4. **Database Expansion**: The method’s computational efficiency could facilitate the expansion of wake-flow databases, providing valuable data for wind farm design, operation, and maintenance.
This article is based on research available at arXiv.